Fulfillment Changes the Commitment, Not Just the Shipment
Most Amazon fulfillment strategies aren't really strategies. They are accumulations.
FBA was the default when the brand started. FBM got added when fees became difficult to absorb on certain products. Another method appeared when a customer or channel needed it. Eventually there are two or three or four methods operating across the same assortment, and nobody can clearly explain why one SKU is in FBA while another is being fulfilled somewhere else.
That is where I want to stop looking at the Amazon account as one channel and start looking at the economics SKU by SKU.
Because the fulfillment method decision is not really a shipping decision. It is an inventory commitment decision wearing shipping-decision clothes. The method you choose determines how much inventory has to be positioned where, on whose fee structure, against which demand pattern, at what working-capital exposure. Same SKU, different method, different inventory commitment behind every unit sold.
That is the argument this piece is about.
The margin you see is not the margin that lands
Every Amazon fulfillment method carries its own fee structure. FBA charges fulfillment and storage on Amazon's schedule, with aged-inventory charges when inventory sits too long. FBM leaves fulfillment costs with the seller, whether those costs sit in its own operation, a 3PL, or carrier contracts. Buy with Prime applies Amazon's Prime service and fulfillment fees to DTC orders, so the economics of your own site change when Amazon becomes part of the fulfillment model. Multi-Channel Fulfillment uses Amazon's fulfillment network to fulfill non-Amazon orders at MCF's own fee rates.
None of that is proprietary knowledge. Amazon publishes the fee schedules. What most brands miss is that the flow-through consequence of these fee structures is different SKU by SKU, and the best fulfillment method for one SKU inside your assortment can be the worst method for the SKU sitting next to it.
What that looks like in a real assortment
I worked with a brand in a physical goods category where the product mix ran heavily to oversize items with a smaller-parcel accessory line alongside.
The brand had negotiated favorable freight rates for the oversize product, because that was the majority of shipping volume and the freight partners priced the contract around it. Those rates did not extend to the small-parcel accessories, which shipped as one-off items on a small-parcel contract the brand had less leverage to negotiate. On the surface P&L, the accessories looked like they should be more profitable. Smaller items. Lower COGS. Higher margin percentage. In the flow-through, they were bleeding margin every time they shipped through the brand's own fulfillment.
Amazon FBA changed that math on the accessories specifically. FBA fees were competitive against the brand's own bad small-parcel rates, and moving the accessories to FBA opened up Prime demand that hadn't existed through the DTC site alone.
The oversize product ran under both FBA and FBM at the same time. FBA's oversize fee structure was punitive on large items, but FBA sold roughly four times the unit volume that FBM did, and the total margin dollars from FBA outweighed the higher per-unit margin FBM preserved. FBM stayed alive alongside FBA because the higher per-unit margin was real for the customers willing to accept the FBM shipping experience, and the brand's own freight contract kept those dollars inside the business.
Same brand. Same Amazon presence. Same SKU running under two Amazon methods at once, plus FBA and FBM assignments on other SKUs, because the SKU-level economics dictated different answers even within a single product.
That is not a shipping decision. That is a decision about how much inventory to position at Amazon versus at the brand's own 3PL, on what fee structure, against which demand curve. The FBA accessories require inventory sitting inside FBA warehouses on Amazon's storage terms. The FBM oversize inventory sits in the brand's own warehouse on the brand's own carrying costs. The working-capital exposure looks different in each case. The stockout risk looks different.
Fulfillment method choice is an inventory commitment choice.
The margin percentage is not the margin dollars
Here is where the analysis gets more interesting.
A fulfillment method can improve the economics of each sale while creating a worse inventory commitment. The reverse can also be true. A method that produces less margin per unit can generate more total margin dollars if the change in fulfillment materially increases the volume the SKU can sell.
That is why I would not make the decision on margin percentage alone.
Initial markup can tell you what margin is built into the selling price, but it does not tell you what the business actually keeps after fulfillment, storage, returns, and other operating costs. More importantly, it does not tell you how much inventory and capital are required to produce the sales.
That distinction matters when comparing fulfillment methods.
Adding FBA to a SKU changes more than the path the order takes to the customer. It changes where inventory has to be positioned, how quickly it needs to be replenished, and what happens when it sits too long. Amazon's FBA model brings its own fulfillment, storage, inventory-performance, and capacity considerations. Buy with Prime requires inventory positioned in Amazon's fulfillment network to support those orders. Multi-Channel Fulfillment uses Amazon's network to serve orders from other channels. FBM keeps the inventory outside that network while putting the fulfillment cost and service requirement back into the brand's own operation.
So the question is not simply,
Which method gives us the highest margin percentage?
It is:
Which fulfillment method produces the best economic outcome for this SKU, given the volume it can generate and the inventory commitment required to support that volume?
That is the reconciliation the fulfillment decision has to hold.
And that is why the fulfillment methods a brand accumulated over time deserve to be looked at again. FBA may have been the right answer when the SKU launched. FBM may have been added later for a legitimate reason. Another method may have solved a specific customer or channel problem.
Those decisions can all have been reasonable when they were made.
The question is whether they still make sense for the assortment today.
What the model has to answer
Before an Amazon fulfillment method gets locked in for a SKU or a category, the operator model has to answer three questions:
What are the flow-through dollars per SKU under each method available?
What is the inventory commitment the method requires, and how does it fit alongside the rest of the assortment's commitments?
Is the additional margin worth the additional commitment?
The output of the model is not a single-method decision. It is a SKU-by-SKU or category-by-category method assignment that reflects where each part of the assortment actually makes money after the inventory commitment is priced in.
You deposit dollars, not percentages.
The question is how many margin dollars the SKU generates, how much inventory and capital it takes to generate them, and whether that return justifies the commitment.
Before you add another method
If your Amazon method mix accumulated rather than got designed, the audit is worth doing before the next method addition compounds the problem.
RiverHouse works marketplace inventory as capital allocation, not fulfillment.
The full frame is here: RiverHouse Marketplace Inventory
If your team cannot answer which SKUs are in the right fulfillment method, why they are there, and what inventory commitment that method requires, that is the conversation to have.